The RNA-seq data and clinical data of 306 CC patients and three normal control samples were downloaded from The Cancer Genome Atlas (TCGA, https://tcga-data.nci.Nih.gov/tcga). The RNA-seq data were used to analyze the expression of FAM83H in CC patients and normal samples. A total of 304 CC patients with detailed clinical information were divided into low or high expression groups according to the median expression of FAM83H. Detailed information on these 304 CC patients is shown in Supplementary Table 1 (only online). The association between the expression of FAM83H and overall survival (OS) was assessed using Kaplan-Meier analysis with log rank test used for comparison. Chi-square (χ²) test was used to analyze the association between FAM83H expression and the clinicopathological features of patients with CC. Univariate and multivariate Cox proportional hazards models were performed to evaluate the prognostic value of FAM83H. In addition, we also analyzed the association between expression of FAM83H and OS of CC patients by searching the Human Protein Atlas (https://www.proteinatlas.org) database and Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/) database. Patients in the Human Protein Atlas were divided into high or low expression groups with the best cut-off point, while patients in the GEPIA database were divided into high or low expression groups based on the median of FAM83H expression. The data download process complied with all laws and regulations.

Human CC cell lines HeLa and SiHa were obtained from the Chinese Academy of Sciences (Shanghai, China). Hacat cells were used as a control. These cells were routinely cultured in RPMI-1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Sijiqing Biotec, Hangzhou, China), 100 U/mL penicillin, and 0.1 mg/mL streptomycin (Sigma Chemical Company, Shanghai, China) at 37℃ with 5% CO₂ in a humidified incubator. Non-adherent cells were washed away after 3 days, and adherent cells were incubated with fresh complete medium. The cells were sub-cultured at the split ratio of 1:2 after reaching confluence.

Two FAM83H-specific small interfering RNAs (si-FAM83H#1: 5′-GAGGAAGUCCAACUACAGG-3′ and si-FAM83H#2: 5′-CUACAGGAUUUAUCGAGCA-3′) were synthesized by GenePharma (Shanghai, China). Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA, USA) was used for transfection. Twenty-four hours after transfection, the HeLa cells were lysed and then subjected to Real-time Quantitative Reverse Transcription PCR (qRT-PCR) to examine the efficiency of inhibition. HeLa cells transfected with scrambled siRNA control (si-con) were regarded as controls.

Total RNA was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) from CC cells transfected with siRNA and the negative controls. The extracted RNA was dissolved using RNAase-free water that was treated with diethyl pyrocarbonate. Next, cDNA was synthesized based on total RNA using PrimeScript RT reagent Kit (TaKaRa, Dalian, China). Afterwards, qRT-PCR was conducted to quantify the expression levels of FAM83H. The primers for this gene were as follows: FAM83H, F: 5′-CTGGACCAGAAGATGCCCCT-3′, and R: 5′-CTGGAGGAGATGGCAGACCT-3′. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an endogenous control (F: 5′-GGAGCGAGATCCCTCCAAAAT-3′, R: 5′-GGCT GTTGTCATACTTCTCATGG-3′). The 2^−ΔΔCT method was utilized to calculate the relative expression levels.

Cell proliferation ability was assessed using a Cell-Counting Kit 8 (CCK-8) assay, as described by the manufacturer. The HeLa cells were cultured in 96-well plates, and then, 10 µL of CCK-8 solution was added to each well at the indicated time points (24, 48, 72, and 96 h) after transfection. Further, cells were cultured for 2 h at 37℃ in a 5% CO₂ incubator. Absorbance was measured at 450 nm.

Colony formation was conducted as described previously. 1213 Specifically, si-FAM83H and si-control cells were cultured in medium supplemented with 10% FBS, and replacing the medium every four days. Then, the cells were incubated at 37℃ in a 5% CO₂ incubator for 1 week. Subsequently, 4% paraformaldehyde was used to fix the colonies, and 0.1% crystal violet (Sigma, St. Louis, MO, USA) was applied to stain the colonies. The images of the colonies were photographed. The number of colonies were quantified using ImageJ software (National Institutes of Health, Bethesda, MD, USA). The number of colonies in the si-con group was defined as 100%.

The effect of FAM83H on the migration of HeLa cells was measured using wound-healing assay kits. Wound healing inserts were added into 24-well plates, and the cells were seeded to either side of the insert after 24 h transfection and cultured overnight to form a monolayer. After that, inserts were discarded when an open wound field in the monolayer of cells was generated. Then, cells continued to be cultured for 24 h under standard conditions. Photographs of wound healing were recorded with a phase-contrast microscope at 0 and 24 h after eliminating the inserts.

The migration assay was carried out using a 24-transwell migration chamber with 8-µm pore filters (BD Biosciences, San Jose, CA, USA). HeLa cells were added into the migration chamber where the bottom chamber was filled with 500 µL of complete medium supplemented with 10% FBS. For the invasion assay, 2×10⁵ cells were seeded in the upper chamber containing an 8-µm pore Matrigel Invasion Chamber (BD Biosciences, San Jose, CA, USA). Serum-free medium was put into the lower chamber in the invasion assay. The migration and invasion chambers were incubated at 37℃ overnight. Thereafter, the migrated or invaded cells on lower surface of the filter were stained with 0.1% crystal violet for 20 min and counted in five microscopic fields for each well.

Cells were lysed using RIPA buffer (Beyotime, Jiangsu, China), and protein concentration was measured with Bio-Rad DC protein assay (Bio-Rad, Hercules, CA, USA). Then, a total of 20 µg of protein from the cell lysate was subjected to sodium dodecylsulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride membranes. Subsequently, membranes were blocked with 5% non-fat milk overnight and incubated with the specific primary antibodies (anti-FAM83H, anti-AKT, anti-p-AKT, anti-PI3K, anti-p-PI3K, and anti-GAPDH). Following washing for 30 min with TBST, secondary antibody was used to incubate the membranes in TBST for 2 h at room temperature. After washing, the interested protein signals were captured using Enhanced Chemiluminescence Plus detection kit (Thermo Fisher Scientific, Rockford, IL, USA).

One-way analysis of variance (ANOVA) was utilized to compare the values of multi-groups. All experiments were carried out in triplicate. Statistical analysis was performed using SPSS 22.0 software (IBM Corp., Armonk, NY, USA). p values <0.05 were considered statistically significant.

By retrieving the data in the TCGA database and GEPIA database, we characterized FAM83H expression in CC (Fig. 1A–C). Results revealed marked up-regulation of FAM83H expression in CC tissues, compared to normal tissues (Fig. 1A–C).

In order to verify the bioinformatics results, we next measured the expression of FAM83H in the CC cell lines SiHa and HeLa using qRT-PCR. Consistent with the bioinformatics results, FAM83H expression was remarkably up-regulated in SiHa and HeLa cell lines when compared with normal Hacat cells (p<0.001) (Fig. 1D). We found that expression levels in HeLa cells were significantly higher than those in SiHa cells. Thus, in our study, we selected HeLa cells for subsequent analysis.

We then analyzed the relationship between FAM83H and clinical pathological characteristics of CC based on the data obtained from TCGA database. The clinicopathological characteristics of CC patients are shown in Table 1. We found that high FAM83H expression in CC was associated with clinical stage (p=0.008), pathologic tumor (p=0.023), and pathologic node (p=0.038).

To further investigate the prognostic value of FAM83H, we performed Kaplan-Meier analysis based on the TCGA, GEPIA, and the Human Protein Atlas databases. The results showed that high FAM83H expression is associated with worse OS in patients with CC based on the TCGA data with the median of FAM83H expression used as the cutoff (Fig. 1E, p=0.027). Although the difference was not significant, we also noted that high FAM83H expression was associated with lower survival rate to some extent in the GEPIA database (Fig. 1F, p=0.053). From the Human Protein Atlas database, we found that higher expression of FAM83H was associated with shorter survival with the best cut-off point (Fig. 1G, p=0.004).

Next, we performed univariate and multivariate proportional hazards analyses to identify independent prognostic factors for survival in CC patients. Univariate analysis suggested that FAM83H expression (p=0.028), clinical stage (p=0.001), pathologic tumor (p<0.001), pathologic metastasis (p=0.020), and pathologic node (p=0.003) were significantly related to the OS of CC patients (Table 2). Multivariate analysis demonstrated that pathologic tumor was an independent prognostic factor for the OS of CC patients (Table 2) (p=0.033). However, FAM83H expression could not be regarded as independent prognostic factor for the OS of CC patients (p=0.716). Above all, the results indicated that, while the expression of FAM83H is significantly associated with OS in CC patients, it cannot be considered an independent prognostic factor.

To identify the roles of FAM83H in CC cells, si-FAM83H was transfected into HeLa cells to determine its effect on the proliferation ability of HeLa cells. In the current study, the transfection efficiency of both si-FAM83H #1 and si-FAM83H #2 was >80% (Fig. 2). In addition, a greater decrease in FAM83H expression was observed in the si-FAM83H #1 group, compared with the si-FAM83H #2. Thus, si-FAM83H #1 was used in subsequent experiments.

CCK-8 results suggested that the knockdown of FAM83H effectively inhibited the proliferation of HeLa cells (Fig. 3) at 48 h (p<0.05), 72 h (p<0.001), and 96 h (p<0.001). Likewise, colony formation assay showed that FAM83H silencing reduced the colony numbers of HeLa cells (p<0.001) (Fig. 4).

Following 24 h scratch formation in the cell monolayer, the wound in the FAM83H knockdown cells was wider, compared with that in the control cells, suggesting that knockdown of FAM83H significantly inhibits HeLa cell migration (Fig. 5A). Subsequently, we also utilized transwell assay to investigate the potential effects of FAM83H knockdown on cell migration and invasion. We observed that, after FAM83H silencing, the migration and invasion of CC cells were effectively inhibited (p<0.001) (Fig. 5B and C). Overall, these data indicated that FAM83H promotes CC cell migration and invasion.

To further describe the specific molecular mechanism of FAM83H-regulated proliferation and migration of CC cells, we used Western blotting to explore the candidate targets of FAM83H, by which we selected the PI3K signaling pathway (AKT, p-AKT, PI3K, p-PI3K) as a target, since the PI3K pathway plays key roles in mediating proliferation and migration. 141516 Fig. 6 shows that knockdown of FAM83H remarkably inhibited the expression levels of p-AKT and p-PI3K in CC cells (p<0.001), but did not influence total PI3K and AKT expression levels.